CURACO-enhanced AI inpatient heart failure tools: a pilot study on clinical confidence and patient-centred care

Abstract Background/Introduction Inpatient heart failure management requires systematic assessment and medication optimisation, yet resident doctors often lack confidence managing these complex patients, potentially leading to suboptimal care. Traditional implementations focus primarily on clinical competency without addressing patient empowerment, healthcare equity, or systematic integration of safety, understanding, research, authenticity, ethics, and technology. Current approaches lack holistic frameworks ensuring clinical excellence and patient-centred care. Purpose To conduct a pilot evaluation of whether AI-generated inpatient heart failure checklists enhanced through CURACO principles (Clinical safety, Understanding, Research-informed care, Authentic patient-centred approaches, Conscientious ethics, and Optimised technology) improve resident confidence while incorporating patient empowerment and equity considerations. Methods We developed AI-generated inpatient heart failure checklists at our hospital, derived from ESC guidelines and enhanced using CURACO principles to ensure safety-first protocols, patient understanding components, research-informed guidelines, authentic care approaches, ethical considerations, and optimised technology integration. The checklists structured evaluation of clinical congestion, medication optimisation focusing on four pillars of heart failure therapy, monitoring parameters, and specialist referral pathways. We surveyed resident doctors (n=10) between August 2024-January 2025 using 5-point Likert scales across five domains before and after implementation. Paired t-tests analysed confidence scores. Results All domains showed significant confidence improvements (p<0.001). High confidence (≥4/5) increased substantially: initial patient assessment (10% to 80%), medication management (0% to 70%), and clinical monitoring (10% to 80%). Similar improvements occurred in diuretics and fluid management (10% to 70%) and specialist referral (20% to 80%). Mean scores improved by 1.7-2.1 points across all care aspects. The CURACO-enhanced approach demonstrated additional benefits: 100% agreed the checklist helped standardise their approach while maintaining patient-centered focus. Patient understanding components were valued by 90% of users, creating a more holistic approach. Conclusion Implementation of CURACO-enhanced AI-generated inpatient heart failure checklists significantly improved resident confidence across all domains while incorporating patient empowerment principles. This pilot demonstrates that AI-derived interventions can be transformed through systematic integration of safety, understanding, research, authenticity, ethics, and technology. The CURACO framework provides foundation for developing clinical tools serving both clinical excellence and patient-centered care. These findings justify progression to larger studies with patient outcomes and multi-centre validation.